Displaying #cassandra-dev/2016-11-18.log:

Fri Nov 18 00:37:47 2016  kohlisankalp:Joined the channel
Fri Nov 18 00:53:36 2016  dikang:Joined the channel
Fri Nov 18 00:57:29 2016  dikang_:Joined the channel
Fri Nov 18 01:14:52 2016  kohlisankalp:Joined the channel
Fri Nov 18 01:17:47 2016  kohlisankalp:Joined the channel
Fri Nov 18 01:57:43 2016  dikang:Joined the channel
Fri Nov 18 02:06:43 2016  vince_:Joined the channel
Fri Nov 18 02:09:01 2016  dikang:Joined the channel
Fri Nov 18 03:01:59 2016  kohlisankalp:Joined the channel
Fri Nov 18 06:09:54 2016  vince_:Joined the channel
Fri Nov 18 08:01:42 2016  kvaster:Joined the channel
Fri Nov 18 08:33:08 2016  spodkowinski:Joined the channel
Fri Nov 18 09:13:05 2016  sergey_dobrodey:Joined the channel
Fri Nov 18 09:14:58 2016  sergey_dobrodey:Hi to all, I would like to draw attention to the problem https://issues.apache.org/jira/browse/CASSANDRA-12916
Fri Nov 18 10:01:18 2016  beobal:sergey_dobrodey: thanks, that looks like a bug that's been present since 3.6. I'll take a look at the patch today
Fri Nov 18 10:04:42 2016  sergey_dobrodey:Real problem is in UserType validate() method. https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/db/marshal/UserType.java#L174 Validation is copied from TupleType but UDT doesn't store all data in one buffer
Fri Nov 18 10:08:30 2016  sergey_dobrodey:My patch from CASSANDRA-12916 doesn't cover situation when UDT is stored inside Collection type.
Fri Nov 18 10:08:31 2016  CassBotJr:https://issues.apache.org/jira/browse/CASSANDRA-12916 (Open; Unresolved; Unscheduled): "Broken UDT muitations loading from CommitLog"
Fri Nov 18 15:14:21 2016  vjarnot:Joined the channel
Fri Nov 18 15:57:37 2016  kohlisankalp:Joined the channel
Fri Nov 18 16:08:48 2016  ostefano:clohfink_, re CASSANDRA-12924, can I safely mix different metrics minor versions? In other words deploying 3.1.2 metrics-graphite, while C* provides metrics-core-3.1.0?
Fri Nov 18 16:08:49 2016  CassBotJr:https://issues.apache.org/jira/browse/CASSANDRA-12924 (Open; Unresolved; Unscheduled): "GraphiteReporter does not reconnect if graphite restarts"
Fri Nov 18 16:09:30 2016  clohfink_:ostefano: Its worth a try, I dont think they broke any interfaces in that
Fri Nov 18 16:09:53 2016  clohfink_:pretty sure it will work, but I am not intimately familiar with the metrics-graphite versions. maybe something to ask on metrics mailing list
Fri Nov 18 16:10:23 2016  ostefano:https://groups.google.com/forum/#!topic/metrics-user/punk1Bx08qc
Fri Nov 18 16:10:25 2016  ostefano::)
Fri Nov 18 16:15:24 2016  thobbs:Joined the channel
Fri Nov 18 17:36:38 2016  iamaleksey:exlt: what's the current state of 3.10? or, rather, when committing a new 3.X bugfix, should I still put it under 3.10 or create a 3.11 section?
Fri Nov 18 17:37:40 2016  exlt:3.10 - current status: blockers still not committed - https://issues.apache.org/jira/issues/?jql=project%20%3D%20CASSANDRA%20AND%20fixVersion%20%3D%203.10%20AND%20resolution%20%3D%20Unresolved%20ORDER%20BY%20due%20ASC%2C%20priority%20DESC%2C%20created%20ASC
Fri Nov 18 17:37:59 2016  iamaleksey:under 3.10 then
Fri Nov 18 17:38:00 2016  iamaleksey:thanks
Fri Nov 18 17:38:06 2016  exlt:yep yep!
Fri Nov 18 17:39:24 2016  beobal:ftr, Pavel committed the revert of 11990 yesterday, so that can have its fixver bumped to 3.x
Fri Nov 18 17:39:37 2016  exlt:I'll be on vacation starting this evening through 11/27, but can probably work in a 3.10 release build/vote, if the status changes
Fri Nov 18 17:39:52 2016  beobal:isn't 11990 the cause of 12877
Fri Nov 18 17:41:21 2016  exlt:ooh, thanks for the revert commit note - I don't see 11990 referenced anywhere on 12877
Fri Nov 18 17:41:39 2016  exlt:I do see the reverse
Fri Nov 18 17:41:57 2016  beobal:yeah
Fri Nov 18 17:42:31 2016  beobal:it's linked the other way, but only in the comments it seems
Fri Nov 18 17:42:32 2016  exlt:so.. both of these are now fixver 3.X?
Fri Nov 18 17:42:55 2016  beobal:well, I guess 12877 can probably be closed, and 11990 bumped to 3.X
Fri Nov 18 17:47:00 2016  exlt:ifesdjeen: could you comment on the above, if you're around? Can 12877 be closed and 11990 set to fixver 3.x?
Fri Nov 18 17:49:06 2016  ifesdjeen:Yes 11990 is the cause. We can. Lose 12877 and re-roll. 11990 will be re implemented without that limitation (in fact already is just need to test it well)
Fri Nov 18 17:52:22 2016  exlt:iamaleksey: just checking, #12281 is the commit you were asking about?
Fri Nov 18 17:52:23 2016  CassBotJr:https://issues.apache.org/jira/browse/CASSANDRA-12281 (Resolved; Fixed; 2.2.9, 3.0.11, 3.10): "Gossip blocks on startup when another node is bootstrapping"
Fri Nov 18 17:52:45 2016  iamaleksey:yup
Fri Nov 18 17:52:50 2016  exlt:awesome, thanks
Fri Nov 18 17:53:15 2016  iamaleksey:it wasn't one of the blockers, was it?
Fri Nov 18 17:53:31 2016  exlt:nope, but that's just fine to be in the release :)
Fri Nov 18 18:14:16 2016  exlt:man, every other jira operation is throwing errors..
Fri Nov 18 18:42:58 2016  kohlisankalp:Joined the channel
Fri Nov 18 18:44:15 2016  vjarnot:Joined the channel
Fri Nov 18 23:09:25 2016  kohlisankalp:Joined the channel
Fri Nov 18 23:32:59 2016  clohfink:Joined the channel

Comments